Erson Analysis Vignette

4. All Custom Functions Used

CreateSeuratObject

Explanation:

More easily creates a Seurat object based on file directory
Parameters:
- fileDirectory: file path (must be h5 file for this specific function)
- objectName: name of the Seurat object itself (e.g. “sample1”)
- projectName: metadata label (e.g. “AsuzuProject”)
  - Used mostly for bookkeeping in Seurat (acts as plot label, etc.)

CreateSeuratObject <- function(fileDirectory, objectName, projectName) {
  rawCounts <- Read10X_h5(fileDirectory)
  seurat_obj <- Seurat::CreateSeuratObject(counts = rawCounts, 
                                           project = projectName, 
                                           min.cells = 3, 
                                           min.features = 200)
  assign(objectName, seurat_obj, envir = .GlobalEnv)
  return(objectName)
}

Example:

CreateSeuratObject("~/Documents/DATA1.H5",
                   "ex_sample1", 
                   "ex_sample1")

AddMitoRna

Explanation:

Appends mitochondrial RNA data to Seurat object on a cell-by-cell basis

AddMitoRna <- function(seuratObject){
  seuratObject[["percent.mt"]] <- Seurat::PercentageFeatureSet(seuratObject, pattern = "^MT-")
  return(seuratObject)
}

Example

ex_sample1 <- AddMitoRna(ex_sample1)

MakeVlnPlot

Explanation:

More easily creates violin plots for QC purposes
- nFeature_RNA = gives the number of unique features (genes) in a given cell, each point represents one cell
- nCount_RNA = gives the total number of RNA molecules in a cell, each point represents one cell
- percent.mt = percentage of mitochondrial RNA, each point represents one cell

MakeVlnPlot <- function(seuratObject) {
  print(VlnPlot(seuratObject, 
                features = c("nFeature_RNA", 
                             "nCount_RNA", 
                             "percent.mt"), 
                ncol = 3))
  return(seuratObject)
}

Example:

# Must be run previously:
CreateSeuratObject("DATA1.H5",
                   "ex_sample1", 
                   "ex_sample1")
ex_sample1 <- AddMitoRna(ex_sample1)

MakeVlnPlot(ex_sample1)

## Warning: Default search for "data" layer in "RNA" assay yielded no results; utilizing "counts" layer instead.

## An object of class Seurat 
## 18197 features across 3301 samples within 1 assay 
## Active assay: RNA (18197 features, 0 variable features)
##  1 layer present: counts

QCSeurat

Explanation:

Streamlines the quality control process, allows for low-quality cells to be eliminated based on feature count and mitochondrial RNA percentage
Parameters usually decided based on violin plots to minimize outliers/maximize viable retained data
Parameters:
- seuratObject = seuratObject to be trimmed
- minFeature = minimum number of genes a cell must contain to remain included in data
- maxFeature = maximum number of genes a cell can contain to remain included in data
- percentMT = maximum percentage of mitochondrial RNA which a cell can contain to remain included in data

QCSeurat <- function(seuratObject, minFeature, maxFeature, percentMT) {
  seuratObject <- subset(seuratObject, subset = nFeature_RNA > minFeature & nFeature_RNA < maxFeature & percent.mt < percentMT)
  return(seuratObject)
}

Example:

# Violin plot prior to QC
MakeVlnPlot(ex_sample2)

## Warning: Default search for "data" layer in "RNA" assay yielded no results; utilizing "counts" layer instead.

## An object of class Seurat 
## 18197 features across 3301 samples within 1 assay 
## Active assay: RNA (18197 features, 0 variable features)
##  1 layer present: counts

ex_sample2 <- QCSeurat(ex_sample2, minFeature = 200, maxFeature = 3000, percentMT = 25)

#Violin plot following QC
MakeVlnPlot(ex_sample2)

## Warning: Default search for "data" layer in "RNA" assay yielded no results; utilizing "counts" layer instead.

## An object of class Seurat 
## 18197 features across 3121 samples within 1 assay 
## Active assay: RNA (18197 features, 0 variable features)
##  1 layer present: counts

NormalizeAndVariable

Explanation:

Normalizes data and finds genes within Seurat object that display variable expression patterns and creates a plot highlighting the top 10 genes with the highest degree of variability in their expression.

NormalizeAndVariable <- function(seuratObject) {
  seuratObject <- NormalizeData(seuratObject)
  seuratObject <- FindVariableFeatures(seuratObject, selection.method = "vst", nfeatures = 2000)
  top10 <- head(VariableFeatures(seuratObject), 5)
  plot1 <- VariableFeaturePlot(seuratObject)
  plot2 <- LabelPoints(plot = plot1, points = top10, repel = TRUE)
  print(plot2)
  return(seuratObject)
}

Example:

ex_sample3 <- NormalizeAndVariable(ex_sample3)

## Normalizing layer: counts

## Finding variable features for layer counts

## When using repel, set xnudge and ynudge to 0 for optimal results

## Warning in scale_x_log10(): log-10 transformation introduced infinite values.

Scale

Explanation:

Scales the Seurat object as per the workflow outlined in the Seurat vignette

Scale <- function(seuratObject) {
  seuratObject <- ScaleData(seuratObject, features = rownames(seuratObject))
  seuratObject <- ScaleData(seuratObject, vars.to.regress = "percent.mt",)
  return(seuratObject)
}

Example

ex_sample4 <- Scale(ex_sample4)

PcaElbow

Explanation:

Runs a principal component analysis and creates an elbow plot based on it

PcaElbow <- function(seuratObject) {
  seuratObject <- RunPCA(seuratObject, features = VariableFeatures(seuratObject))
  print(ElbowPlot(object = seuratObject))
  return(seuratObject)
}

Example:

# Must be run previously
ex_sample4 <- Scale(ex_sample4)

ex_sample4 <- PcaElbow(ex_sample4)

## PC_ 1 
## Positive:  IL1B, LYZ, PLAUR, CTSS, IER3, CXCL8, HLA-DRA, FCN1, OLR1, SRGN 
##     S100A9, VCAN, BCL2A1, TYROBP, AIF1, EREG, FCER1G, NLRP3, SOD2, CXCL2 
##     CD74, FGL2, CCL20, S100A8, MS4A7, VIM, THBS1, G0S2, MNDA, NFKBIA 
## Negative:  CHGB, GNAS, SCG2, MEG3, PEG10, NNAT, SEC11C, MIR7-3HG, CCND1, RPS24 
##     CALY, RAB3B, SHTN1, HSPA12A, KRT18, TMEM136, CITED1, DLK1, ASCL1, PTN 
##     PEG3, POLR3A, PRDX4, KRT8, ARG2, ORAOV1, CYFIP2, POU1F1, MAP1B, INSM1 
## PC_ 2 
## Positive:  TIMP3, SPARCL1, IGFBP4, IGFBP5, GNG11, IFITM3, A2M, IFI27, PLVAP, SPARC 
##     ABCG2, ADGRF5, KDR, EMCN, RAMP2, CAVIN2, SLCO2A1, PLPP3, STC1, ADGRL4 
##     IGFBP7, EPAS1, PLAT, HYAL2, PCAT19, CD9, IGFBP6, ESAM, PODXL, TGM2 
## Negative:  LYZ, IL1B, CTSS, BCL2A1, S100A9, TYROBP, FCN1, S100A8, EREG, SAMSN1 
##     FCER1G, AIF1, PTPRC, CD44, OLR1, NLRP3, CCL20, S100A12, MNDA, VCAN 
##     LCP1, PLAUR, C5AR1, FGL2, MS4A7, G0S2, CLEC7A, AC020656.1, CSTA, CYBB 
## PC_ 3 
## Positive:  CCND1, SHTN1, RAB3B, RPS24, HSPA12A, TMEM136, ASCL1, INSR, ORAOV1, PMAIP1 
##     SPP1, POLR3A, TBX19, RCN1, ARG2, PCSK1, SCG2, GAL, GADD45B, PPP1R17 
##     CYP3A4, RGS16, CREG1, INSM1, FILIP1, PRDX4, CHGB, TRDN, OSBPL1A, NEB 
## Negative:  TMSB4X, NKG7, B2M, CCL5, IL32, KLRB1, GZMA, CD52, GZMB, CST7 
##     GNLY, CD3D, CXCR4, HCST, TRBC2, S100A4, GZMH, ARL4C, CD2, KLRD1 
##     ZFP36L2, CD7, FGFBP2, CTSW, IL7R, PTPRC, TRBC1, CD247, GZMM, S100A6 
## PC_ 4 
## Positive:  ABCG2, PLVAP, KDR, SLCO2A1, IFI27, EMCN, RAMP2, ADGRL4, CA4, HYAL2 
##     PCAT19, PODXL, HLA-B, ITM2A, TMEM88, ESAM, DNASE1L3, PTPRB, CLEC14A, ITGA6 
##     RGCC, CD34, PLPP3, EGFL7, TM4SF18, ROBO4, MMRN2, SLC9A3R2, CYP26B1, NRP2 
## Negative:  DCN, MGP, BGN, RGS5, TAGLN, FBLN1, IGFBP7, IGFBP2, CCL2, STEAP4 
##     MYL9, CRISPLD2, SERPING1, COL1A2, C1S, GPC3, LUM, NTRK3, NR2F1, APOD 
##     APOE, TPM2, TWIST1, IGFBP6, CALD1, C1R, LAMC3, PCOLCE, RARRES2, COL6A1 
## PC_ 5 
## Positive:  CXCR4, TMSB4X, NKG7, TMSB10, CCL5, RPS24, IL32, GNLY, CD52, GZMA 
##     GZMB, HLA-B, CST7, KLRB1, CD3D, HCST, ZFP36L2, RAB3B, GZMH, S100A4 
##     TRBC2, PTPRC, CD2, ARL4C, CTSW, CD7, CCND1, FGFBP2, KLRD1, ASCL1 
## Negative:  MEG3, DLK1, LINC00632, MIR7-3HG, POU1F1, HTATSF1, ARC, EGR3, GNAS, NNAT 
##     EGR1, SCGN, ARHGAP36, PITX1, CITED1, CADM1, TCEAL5, TAGLN3, ALDH1A1, PEG3 
##     TCEAL6, NR4A1, FOSB, C14orf132, CADPS, EGR4, GH1, VMP1, PRL, ENPP1

MakeHeatmap

Explanation

Creates one heatmap for every principal component included in PcNumberVector
- PcNumberVector should be a numeric vector

MakeHeatmap <- function(seuratObject, PcNumberVector){
  DimHeatmap(seuratObject, dims = PcNumberVector, balanced = TRUE)
  return(seuratObject)
}

Example:

# Note: PcaElbow must be run first, as it attaches PCA data to the Seurat object
MakeHeatmap(ex_sample5, 1:9)

## An object of class Seurat 
## 18197 features across 3121 samples within 1 assay 
## Active assay: RNA (18197 features, 2000 variable features)
##  3 layers present: counts, data, scale.data
##  1 dimensional reduction calculated: pca

FindCellClusters

Explanation:

Finds clusters of cells based off principal component behavior
Parameters:
- seuratObject = self-explanatory
- PCs = a numeric vector containing the principal components to be used
- resolution = a value between 0 and 1; defines the granularity of the clusters
Cluster assignments may be found via the dataframe seuratObject@meta.data, column = seurat_clusters

FindCellClusters <- function(seuratObject, PCs, resolution) {
  seuratObject2 <- FindNeighbors(seuratObject, dims = PCs)
  seuratObject3 <- FindClusters(seuratObject2, resolution = resolution)
  return(seuratObject3)
}

Example:

# Below code finds 11 clusters
ex_sample6 <- FindCellClusters(ex_sample6, 1:7, 0.5)

## Computing nearest neighbor graph

## Computing SNN

## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
## 
## Number of nodes: 3121
## Number of edges: 91787
## 
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.9101
## Number of communities: 13
## Elapsed time: 0 seconds

MakeUMAP

Explanation:

Runs UMAP + creates a UMAP plot based on the specified principal components
Parameters:
- seuratObject = self-explanatory
- PCs = a numeric vector containing the principal components to be used

MakeUMAP <- function(seuratObject1, PCs) {
  seuratObject2 <- RunUMAP(seuratObject1, dims = PCs)
  print(DimPlot(seuratObject2, reduction = "umap"))
  return(seuratObject2)
}

Example:

ex_sample7 <- MakeUMAP(ex_sample7, 1:7)

## 08:26:17 UMAP embedding parameters a = 0.9922 b = 1.112

## 08:26:17 Read 3121 rows and found 7 numeric columns

## 08:26:17 Using Annoy for neighbor search, n_neighbors = 30

## 08:26:17 Building Annoy index with metric = cosine, n_trees = 50

## 0%   10   20   30   40   50   60   70   80   90   100%

## [----|----|----|----|----|----|----|----|----|----|

## **************************************************|
## 08:26:18 Writing NN index file to temp file /var/folders/xs/84n7_bk52l98qmnzc7ndzlmr0000gn/T//RtmphMbDkA/file2ce43adf1801
## 08:26:18 Searching Annoy index using 1 thread, search_k = 3000
## 08:26:18 Annoy recall = 100%
## 08:26:19 Commencing smooth kNN distance calibration using 1 thread with target n_neighbors = 30
## 08:26:20 Initializing from normalized Laplacian + noise (using RSpectra)
## 08:26:20 Commencing optimization for 500 epochs, with 123504 positive edges
## 08:26:22 Optimization finished

FindPosMarkers

Explanation:

Finds markers that are upregulated within previously established clusters

FindPosMarkers <- function(seuratObject){
  posMarkers <- FindAllMarkers(seuratObject, only.pos = TRUE)
  posMarkers %>%
    group_by(cluster) %>%
    dplyr::filter(avg_log2FC > 1)
  print(posMarkers)
  return(seuratObject)
}

Example:

FindPosMarkers(ex_sample8)

Top5Markers

Explanation:

Cleaner version of FindPosMarkers
Prints out the top 5 most upregulated genetic markers on a cluster-by-cluster basis

Top5Markers <- function(seuratObject) {
  posMarkers <- FindAllMarkers(seuratObject, only.pos = TRUE)
  separatedMarkers <- posMarkers %>%
    dplyr::filter(avg_log2FC >1) %>%
    group_by(cluster) %>%
    slice_head(n = 5)
  print(separatedMarkers, n = Inf)
}

Example:

# the actual output is 65 rows
head(Top5Markers(ex_sample8))

SingleHPCA

Explanation:

Annotates each individual cell based on its expression profile’s closest match within the Human Primary Cell Atlas reference dataset; based on the use of SingleR
- HPCA dataset is more generic than the below neuroendocrine or immune cell datasets, but is good for establishing an initial tissue composition overview
- Cell identities may be found via the dataframe seuratObject@meta.data, column = HPCACellLabels

SingleHPCA <- function(seuratObject){
  HpcaLabels <- SingleR(test = as.SingleCellExperiment(seuratObject), 
                        ref =    HumanPrimaryCellAtlasData(), 
                        labels = HumanPrimaryCellAtlasData()$label.main)
  seuratObject1 <- AddMetaData(seuratObject, 
                               metadata = HpcaLabels@listData[["labels"]], 
                               col.name = "HPCACellLabels")
  return(seuratObject1)
}

Example:

ex_sample9 <- SingleHPCA(ex_sample9)

Idents(ex_sample9) <- ex_sample9@meta.data$HPCACellLabels
DimPlot(ex_sample9, reduction = "umap")

SingleNeuro

Explanation: Annotates each individual cell its expression profile’s closest match within the dataset linked from the paper linked here, via the use of SingleR
- Reference dataset is of neuroendocrine cells
- Reference dataset file is in the GitHub repository and should automatically be utilized when calling this function

Reference Dataset

#Just run entire chunk

NeuroMeta <- read.csv("Reference_Datasets/NeuroMeta.csv.gz")

NeuroExpression <- read.csv("Reference_Datasets/NeuroExpression.csv.gz")

columns_to_keep <- colnames(NeuroExpression) %in% NeuroMeta$X | colnames(NeuroExpression) == "X"

NeuroExpressionFiltered <- NeuroExpression[, columns_to_keep]

rm(NeuroExpression)

rownames(NeuroExpressionFiltered) <- NeuroExpressionFiltered[,1]
NeuroExpressionFiltered <- NeuroExpressionFiltered[,-1]

NeuroExpressionFiltered <- NeuroExpressionFiltered %>%
  mutate(across(everything(), as.numeric))

NeuroIDs <- as.character(NeuroMeta$cell_type)
names(NeuroIDs) <- colnames(NeuroExpressionFiltered)

rm(NeuroMeta)
NeuroExpressionFiltered <- as.matrix(NeuroExpressionFiltered)

Function

SingleNeuro <- function(seuratObject) {
  # Function to run SingleR
  NeuroClusterLabels <- SingleR(test = as.SingleCellExperiment(seuratObject),
                                ref = NeuroExpressionFiltered,
                                labels = NeuroIDs)
  
  #To add SingleR output to Seurat Object
  seuratObject1 <- AddMetaData(seuratObject, metadata = NeuroClusterLabels@listData[["labels"]], col.name = "NeuroCellLabels")
  
  #To group everything by Neuro Cell Type
  Idents(seuratObject1) <- "NeuroCellLabels"
  return(seuratObject1)
}

Example:

ex_sample10 <- SingleNeuro(ex_sample10)

DimPlot(ex_sample10, 
        reduction = "umap",
        group.by = "NeuroCellLabels")

SingleImmune

Explanation:

Annotates each cell identified based on its expression profiles’ closest match within the dataset from the paper linked here, via the use of SCINA.
Cell identities may be found via the dataframe seuratObject@meta.data, column = ImmuneCellLabels
Note: Notice that some of the cells from the previous SingleNeuro() function were labeled “Imm”. These are typically the only cells that I ran this function on, so as to not mischaracterize previously identified neuroendocrine cells.
Note: SCINA is used in place of SingleR due to the different structure of the reference dataset.
Note: The function is named SingleImmune purely for syntax consistency, as the function serves the same effective purpose as the aforementioned SingleHPCA despite using SCINA instead of SingleR.
- Reference dataset is of immune cells
- Reference dataset file is in the GitHub repository and should automatically be utilized when calling this function

Reference Dataset

ImmuneRef <- read.csv("Reference_Datasets/ImmuneRef.csv")
ImmuneRef_list <- lapply(1:39, function(i) {
  genes <- ImmuneRef[, i]
  genes <- genes[!is.na(genes)]
  return(genes)
})

names(ImmuneRef_list) <- colnames(ImmuneRef)
ImmuneRef <- ImmuneRef_list
rm(ImmuneRef_list)

Function

SingleImmune <- function(seuratObject) {
  
SeuratObjectExpressionData <- as.matrix(GetAssayData(seuratObject, assay = "RNA", layer = "data"))

#NOTE: significant overlap of genetic signatures between cell types--had to set rm_overlap to FALSE

results <- SCINA(SeuratObjectExpressionData, ImmuneRef, rm_overlap = FALSE)
seuratObject1 <- AddMetaData(seuratObject, metadata = results$cell_labels, col.name = "ImmuneCellLabels")
Idents(seuratObject1) <- "ImmuneCellLabels"
return(seuratObject1)
}

Example:

#notice that we are leveraging our SingleNeuro characterizations
ex_sample10 <- SingleImmune(ex_sample10)
ex_sample10imm <- subset(ex_sample10, NeuroCellLabels == "Imm")

# this dataset results in a VERY large number of cell IDs, so for legibility, we are getting rid of any cell ID with less than 5 cells

CategoryCounts <- ex_sample10imm@meta.data %>%
  group_by(ImmuneCellLabels) %>%
  tally()

ValidCategories <- CategoryCounts %>% 
  filter(n >= 10) %>%
  pull(ImmuneCellLabels)

DimPlot(subset(ex_sample10imm,  ImmuneCellLabels %in% ValidCategories), 
        reduction = "umap", 
        group.by = "ImmuneCellLabels")

SingleUnbiasedNeuro

Explanation: Annotates each individual cluster previously identified via principal component analysis (aka via FindCellClusters) based on its collective expression profile’s closest match within the dataset linked from the paper linked here, via the use of SingleR
- Reference dataset is of neuroendocrine cells (same as SingleNeuro)
- Reference dataset file is in the GitHub repository and should automatically be utilized when calling this function
- Annotations can be referenced via seuratObject@meta.data$UnbiasedNeuroCellLabels

SingleUnbiasedNeuro <- function(seuratObject){
  
# Create a table of seurat_clusters vs NeuroCellLabels
  cluster_label_counts <- table(seuratObject@meta.data$seurat_clusters, 
                                seuratObject@meta.data$NeuroCellLabels)
  
# Identify the most abundant label for each cluster
  most_abundant_labels <- apply(cluster_label_counts, 1, function(x) {
    names(x)[which.max(x)] # Returns the label with the highest count
  })
  
# Relabel seurat_clusters with the most abundant NeuroCellLabels
  seuratObject@meta.data$UnbiasedNeuroCellLabels <- factor(seuratObject@meta.data$seurat_clusters, levels = names(most_abundant_labels), labels = most_abundant_labels)
  return(seuratObject)
}

Example:

ex_sample10 <- SingleUnbiasedNeuro(ex_sample10)

# notice that it is each cluster within "seurat_clusters" that is relabeled 
DimPlot(ex_sample10, 
        reduction = "umap",
        group.by = "seurat_clusters")

DimPlot(ex_sample10, 
        reduction = "umap",
        group.by = "UnbiasedNeuroCellLabels")

SingleUnbiasedImmune

Explanation: Annotates each individual cluster previously identified via principle component analysis (aka via FindCellClusters) based on its expression profiles’ closest match within the dataset from the paper linked here, via the use of SCINA.
Note: SCINA is used in place of SingleR due to the different structure of the reference dataset.
Note: The function is named SingleUnbiasedImmune purely for syntax consistency, as the function serves the same effective purpose as the aforementioned SingleUnbiasedNeuro, despite using SCINA instead of SingleR.
- Reference dataset is of immune cells
- Reference dataset file is in the GitHub repository and should automatically be utilized when calling this function

SingleUnbiasedImmune <- function(seuratObject){
  cluster_label_counts <- table(seuratObject@meta.data$seurat_clusters, 
                                seuratObject@meta.data$ImmuneCellLabels)
  
  most_abundant_labels <- apply(cluster_label_counts, 1, function(x) {
    names(x)[which.max(x)] 
  })
  
  seuratObject@meta.data$UnbiasedImmuneCellLabels <- factor(seuratObject@meta.data$seurat_clusters, levels = names(most_abundant_labels), labels = most_abundant_labels)
  
  return(seuratObject)
}

Example:

# notice that we are using the Seurat object that is subsetted to only have immune cells (identified via SingleNeuro)
ex_sample10imm <- SingleUnbiasedImmune(ex_sample10imm)

# notice that it is each cluster within "seurat_clusters" that is relabeled 
DimPlot(ex_sample10imm, 
        reduction = "umap",
        group.by = "seurat_clusters")

DimPlot(ex_sample10imm, 
        reduction = "umap",
        group.by = "UnbiasedImmuneCellLabels")

BarPlotIdPercent

Explanation: Creates a barplot detailing the percentage each cell identity makes up of the total sample
- Note: the Identity parameter needs to be a character string in quotation marks for the function to work properly
Parameters:
- Identity: the cell annotation that will be used to make the barplot
  - e.g. "NeuroCellLabels" or "seurat_clusters"

#To create an inclusive barplot based on cell identities 
BarPlotIdPercent <- function(seuratObject, Identity){
  
  TotalCells <- dim(seuratObject@meta.data)[1]
  
  cellIdentities <- seuratObject@meta.data[[Identity]]
  
  cellCounts <- table(cellIdentities)
  
  cellCountsDF <- as.data.frame(cellCounts)
  
  colnames(cellCountsDF) <- c("CellIdentity", "Count")
  
  ggplot(cellCountsDF, aes(x = CellIdentity, 
                           y = Count/TotalCells*100, 
                           fill = CellIdentity)) + 
    geom_bar(stat = "identity") +
    theme_minimal() + 
    labs(title = "Distribution of Cell Types", x = "Cell Identity", y = "Percentage of Cells") +
    theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
    theme(text = element_text("sans"))
}

Example:

BarPlotIdPercent(ex_sample10, "NeuroCellLabels")

Erson Analysis Vignette

Sebastian Reyes

2025-06-5

0. Knitting this Document

1. Introduction

2. Installing Packages Needed

3. Loading Packages Needed

4. All Custom Functions Used

CreateSeuratObject

Explanation:

Example:

AddMitoRna

Explanation:

Example

MakeVlnPlot

Explanation:

Example:

QCSeurat

Explanation:

Example:

NormalizeAndVariable

Explanation:

Example:

Scale

Explanation:

Example

PcaElbow

Explanation:

Example:

MakeHeatmap

Explanation

Example:

FindCellClusters

Explanation:

Example:

MakeUMAP

Explanation:

Example:

FindPosMarkers

Explanation:

Example:

Top5Markers

Explanation:

Example:

SingleHPCA

Explanation:

Example:

SingleNeuro

Reference Dataset

Function

Example:

SingleImmune

Explanation:

Reference Dataset

Function

Example:

SingleUnbiasedNeuro

Example:

SingleUnbiasedImmune

Example:

BarPlotIdPercent

Example: